Welcome!

This paper studies how a privately informed sender communicates through an intermediary that classifies the sender into discrete categories (e.g., ratings, tiers, or labels). This discrete structure can restrain strategic misreporting, and accurate implementation of a classification standard is possible only when it uses a small number of categories. Pursuing accuracy, however, comes at a cost, since finer classifications may convey more information by tolerating some misclassification. In a general design problem with an endogenous standard, every classification pools a nondegenerate interval of top types, which places an upper bound on the informativeness attainable for any classification.

We develop a model where firms choose whether to comply with an accounting rule that places an upper bound, such as historical cost, on the report and mandates recognition of bad news. Firms may also make unregulated assertions that their value exceeds the upper bound permitted by the accounting rule. In this context, improved enforcement of accounting rules enhances the credibility of firms’ unregulated claims. Consistent with the model, we use difference-in-differences (DiD) analyses and find that the market reacts more strongly to management guidance after the critical audit matter (CAM) disclosure requirement. The model also shows that firms’ willingness to recognize large impairments depends mostly on misreporting costs, while their willingness to recognize small impairments depends mostly on investors’ prior beliefs and whether a small impairment would be interpreted favorably. Using this feature, we conduct a structural estimation and show substantial noncompliance with goodwill impairment requirements among U.S. firms over the past two decades.

Financial statements reflect not only how firms report outcomes, but also the underlying operating, investing, and financing decisions that generate them. We develop a dynamic model in which firms jointly manage operating capacity, operational risk, and liquidity needs through acquisitions, divestitures, search for favorable transaction opportunities, and risk-management activities. We find that firms make capacity adjustments around favorable transaction opportunities generated through costly search. Whether firms use such opportunities to expand or divest operating assets depends on their liquidity, defined as cash relative to operating assets. We show that search effort and deal frequency are U-shaped in liquidity: highest when firms have either low or high liquidity. We further show that operational risk management and capacity-related search move together when liquidity is low but move apart when liquidity is high. The model generates empirical predictions for acquisitions and divestitures, operating cash-flow volatility, payout timing, refinancing activity, and cash-flow patterns associated with different strategic and life-cycle regimes.

Free riding effects discourage and can even prevent investors from exerting efforts to influence firms, i.e., using ‘voice’ in corporate governance. An intermediary, acting on behalf of investors, can overcome such free-rider problems. We show that, absent additional market frictions, the free-riding problem persists, as rational investors choose not to delegate to an intermediary who passes on the costs of their efforts. Because of this, intermediaries benefit from mechanisms that allow them to economize on costly influence efforts, such as interests aligned with managers, negative externalities across portfolio firms, and co-investing with similarly-interested investors. Our results help explain empirical observations related to asset managers’ activities, e.g., around voting behavior, investment screens, diversification, and co-investment.